bare a certain portion, which forms part of the wall, without
projecting. This living spot in the wall is the fountain where the
supply of moisture is renewed. When its reservoir is exhausted by the
conversion of dry dust into mud the miner descends to its chamber,
thrusts its proboscis into the root, and drinks deep from the vat built
into the wall. Its organs well filled, it re-ascends. It resumes work,
damping the hard soil the better to remove it with its talons, reducing
the debris to mud, in order to pack it tightly around it and obtain a
free passage. In this manner the shaft is driven upwards; logic and the
facts of the case, in the absence of direct observation, justify the
assertion.

If the root were to fail, and the reservoir of the intestine were
exhausted, what would happen? The following experiment will inform us: a
larva is caught as it leaves the earth. I place it at the bottom of a
test-tube, and cover it with a column of dry earth, which is rather
lightly packed. This column is about six inches in height. The larva has
just left an excavation three times as deep, made in soil of the same
kind, but offering a far greater resistance. Buried under this short
column of powdery earth, will it be able to gain the surface? If its
strength hold out the issue should be certain; having but lately made
its way through the hard earth, this obstacle should be easily removed.

But I am not so sure. In removing the stopper which divided it from the
outside world, the larva has expended its final store of liquid. The
cistern is dry, and in default of a living root there is no means of
replenishing it. My suspicions are well founded. For three days the
prisoner struggles desperately, but cannot ascend by so much as an inch.
It is impossible to fix the material removed in the absence of
moisture; as soon as it is thrust aside it slips back again. The labour